ebook img

Observation and prediction of harmful algal blooms PDF

41 Pages·2008·9.2 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Observation and prediction of harmful algal blooms

1 Observation and prediction ofharmful algal blooms J.J. Cullen I.I INTRODUCTION Phytoplankton are the principal source offood for life in the sea, and the dynamics ofphytoplankwn communitiesare centrally important to the structureand function ofpelagic ecosystems. Life forms ofphyroplankwn have evolved to exploit different regimesofturbulenceandhencenutrients(Margalef, 1978) hutbytheirnature,plank IOnic food websare struClured by grazing and otherbiological interactions (Ki0rhoe, 1993; Smeracek, 1998). Consequently, physical processes determine the StruCture of pelagic ecosystems, directly by [heir influence on the growth ofphytoplankton and indirectlybyaffecting food-web interactions (Cullen etaI., 2002). Phytoplanktondy namics are particularlyvariable in coastal ecosystems because physical, chemical and biological influences ate forced in complex ways; flows ofwater are constrained by coastlines and shallow, highly variable bathymetry; nutrients are supplied from ter restrial andbenthicsourcesaswell as from deeperwateroffshore; andbenthicand in tertidal communitiesactively interacrwith (hose in (hewatercolumn. The increasing concentrationofhumanactivitiesin(hecoastal region(NichollsandSmall,2002)and the multifaceted importance ofcoastal ecosystems to the sustainability ofthe planet (Costanza etaI., 1997) ensure the prominenceofbiological variability in coastal eco systemsasanenvironmental concern. Transient proliferations of phytoplankton, referted to as blooms, are common and natural in coastal environments. In pelagicsystems, such outbreaks are the prin cipal means bywhich Rowsofmanerand energyescape the tightlycoupled microbial loop to feed higher trophic levels and exportorganic maner to deeperwatersand the bottom assinkingparticles (Michaelsand Silver, 1988; Legendreand Lc Fevre, 1989; Ki0rboe, 1993).Algal bloomsarcrhusintegral to planktonicecosystem dynamicsand biogeochemical cycles. However, some phyroplankton blooms in coastal or brackish watersareperceivedasharmful.Theycancausemassivefish kills,contaminateseafood with toxins, and alterecosystems in ways thathumans do not like. Thesearcharmful algal blooms (HABs), ageneric term thatglossesoverthatfacts that notall HABspe ciesareclassified asalgae and somespecies causeharmful effectswhen present in low cell densities, norhing like a bloom (Smayda, 1997b). Arhorough review ofrhe HAB problemwasrecentlypreparedaspartoftheSciencePlan for an imernational research programmeon HABs (GEOHAB, 2001). As thischaptershares manyobjectiveswith theGEOHAB Science Plan,some material from thatdocument is repeated here,with added emphasis on observation and prediction ofHABs in the context ofreal-time observation systems. ~- ~'-- D- l..3iI- -- .. , ,.- rm::;·-~ _ _- D· -.. ·C2l._- ·..- .- Figure 1.1 Mapsofrep(.medoccurrencesofHABtoxicitydemonstratemanyaspects oftheHABproblem.Topleft:globaldistribUlionsofreponedJnralyrk shdlfishpoisoning(PSP)ariddi:mheticshellfishpoisoning(DSP)events showthatharmfuleffectsarcwidespread.R~ionaJ patternsaretosome extentrelated10thepresenceofsystemsformonitoringandreponing; absenceofreportsdoesnotnecessarilymeanthatHABshavenotoccurred (USNadonalOfficeforMarineBioloxinsandHarmfulAlgal Blooms). Otherpanels:OccurrencesofPSP, DSPandamnesicshellfishpoisoning (ASP) inICESCOUntrieslfrom 1993-2002showthatsomelocationsare affectedbyseveraltypesofHAB,whereascertainIypesoftoxicityarc: morerestricted. Noneofthesemapscanshowthediversityofspeciesthar generatetheseeffeCls(fable LJ).TheCoastalModuleoftheGlobalOcean ObservingSystem (lOc.2003)isbeingestablishedtofacilitatemOte effcctivemonitoringofHABsandOIherphenomenaincoastalenvironmenlS worldwide,solemporalandspatialpanernscanberesolved.andexplained.' Sourct: HarmfulAlgaeEvenl Dala Base(HAEDAT),©IFREMER. 'ICES (International Council for the E~ploration ofthe Sea) membercountries arc Belgium, Can:KIa. Denmark, Estonia, Finland, France,Germany, ledand, Ireland, Latvia, Nerherlands, Norway, Poland, Porcugal,Russian Federation.Spain,Sweden,UnitedKingdomand UnitedS'ates. Jh.tp:/lwww.ifremer.frlenvlitldocumen.alionldossie.slciem/aciem-c!.h,m 2 ObSl:rvarionandprl:dil:.:tionofharmfulalgal blooms l.1.1 DiversityofHABs ThegreatdiversityofHABspeciesandeffects(Table 1.1,Figure1.1) precludeseffective generalization. Still, harmful algaearecommonlyclassified in twOgroups: • toxin producers, which can contaminate seafood, kill fish, orcause health prob lems in humansthrough directexposure tothetoxins; high-biomass producers, which can kill or damage marine life after reaching denseconcentrations, for examplebycausinganoxia aftercollapseofabloom or bychronicallyshadingbenthicvegetation. Bloomsofhigh-biomass producers also affect tourism and recreation bydiscolouring coastal waters and generating noxious foams, slimesor odours. Even this broad clas sificationofharmfulalgaeisnotexclusive: several HABspeciesthataretoxicalsoform denseblooms. To compromisegeneralization further, closely relatedspeciescanoccur in highconcentrations in some regions butnotothers. Forexample,A!~xdndriumdis colourswaterin the GulfofStLawrencebutgenerallyforms only low-densityblooms in theGulfofMaine, although itcontaminatesshellfish ineach environment (Ander son, 1997;Weiseet al., 2002). Otherspecies exhibit variable toxicity, for reasons that are as yet unresolved (e.g. Graneli et al., 1993~ Scholin eta!" 2007- Chapter II this volume). Formal definitions ofalgal blooms are thusneitherpractical norparticularly helpful (Smayda, 1997b);in thischapter, bloom issynonymouswith an increaseinthe abundance ofa phytoplankton species above a background concentration, in either spaceor time. Considered broadly, HABs are agrab-bag ofphenomena with little in common except for effects that humans perceive as being harmful. Due to their diversity, no TAStE 1.1 Somedeleteriouseffectscausedbyharmfulalgaeincoastaland brackishwaters Humanh~alrh Paralyticshellfish poisoning(PSP) Dinoflagellatl:S AI~xandr;umspp" Pyrodinium hahammuvar. compr~ssum, Gymnodiniumcaunatum Cyanobacteria Anahamacircinalil Diatrhericshellfishpoisoning Dinoflagellatl:s Dinophysuspp., Prorountrumspp. (DSP) Neurotoxicshellfishpoisoning Dinoflagellates Karmia hr~vis INSP) Amnesicshellfish poisoning DialOms Puudo-nitzschiaspp. (ASP) Anspiracidshl:llfishpoisoning Dinoflagellate Protop~ridinium aassipa' (AZP) Ciguatl:rafish poisoning(CFP) Dinoflagellates GamhiadilcustOX;CUI Rl:spiratoryproblemsandskin DinoflagellateS Karmia hr~vil, Pfi~suriapilcicida irritation, nl:urologicaleffo:cts Cyanobactl:ria NodulariaIpumigma Hepatotoxicity Cyanobacteria Microcystisa~ruginosa, Nodularia spumigma (Continud) 3 Real-dmeCoa51al ObservingSysrems TABLE I I (continu~d) Examplesofcausaliveorganisms Nalumland(l<llUr~dmarin~rnOllrUI Hemolytic, hepatotoxic, Dinoflagellates Gymnodiniumspp.. Cochlodinium osmoregulatorydTectsandother polykrikoid~s.Hfurocapsa unspeci~edtoxicity circularilquama, Pjifstaiapiscicida. Gonyaulaxspp. RaphidophYles Httaosigmaakalhiwo, Charront!!a spp.,Fibrocapsajaponiea Prymncsiophytcs Chrysochromlliinaspp., Phatoryltis poucherii, Prymnrsiumspp. Cyanobacleria MicrorysriIaauginosa, Nodulariaspp. Negativeeffecrson fuding Pel3gophyres Aurrotoccusanophagrffirrns behaviour Hypoxia,anoxia Dinoflagellates Prorountrummicans, Crratiumfirrca Mechanicaldamage DialOms Chatfocrrosspp. Gillcloggingand necrosis Prymnesiophyles Phatorysrisspp. TourismandrurtillionalaclivitifS Productionoffoam, mucilage, Dinoflagellates Noetilucascinri!!ans. Prorocrntrumspp. discoloration, repellentodour Prymnesiophyles Phaeorystisspp. Diatoms Cylindr(Jfhtcadosrrrium CY3nobacieria Nodulariaspumigena.Aphanizomrnon fios-aquat, Mierorystisatruginosa, Lyngbyaspp. Marinftcosystfmimpacts Hypoxia,anoxia Dinoflagellates Noetilucascintillans, Httrrocapsa triquerra Diatoms SkfltlOntmacosMtum Prymnesiophyle!i Pharoryslisspp. Negativeeffecrson feeding Pe!agophytes AllrtococcusIfnophagtffir~ns, behaviourand reductionofwater AurroumbralagunrnsiI clarily Dinoflagellates Prorountrumminimum Toxiciry10 marineorg3llisms, Dinoflagellates Kareniabrrvis,A/rxandrillmspp. includinginvenebrares, fish, Diatoms Psrudo-nitzschiaI1mtralis mammalsandbirds Sourees: Zingoneand Enevoldscn (2000) asmodified byGEOHAB(2001),with further modifications. 'Recentlydescribed byJamesetal. (2003). one cause for HABs can be found, and no single sUategy for deteerion or prediction willsuffice.Themajorehallengefor understandingHABs[0suPPOrtmanagementand mitigarion is to describe, for each species, wh3t conditions promore its developmem insteadof(or in concenwith) orher phytoplankton (GEOHAB, 2001). 4 Observationandpredictionofharmfulalgalblooms 1.1.2 Harmfulalgaeandenvironmentalvariability ItisaxiomaticthatHABspecieshaveadaptedto manyniches (suitesofecologicalfac tors that determine their distributions and activities), and that the matches between the adaptations ofharmful species and oceanographicvariabilityare good enough to ensuretheirsurvivalfromyeartoyear,andtheirproliferationwhenconditionsarecon ducive.ThegreatdiversityofHABs- with respect to taxonomy, region, hydrographic regimeandharmfuleffects- reflectsinafundamentalwaythatenvironmentalforcing hasselectedforawidevarietyofharmfulalgae,ontimescalesfromevolutionarytodays and spatial scales from ocean basins to bays. For each harmful species, the challenge is to unlockits secrets: whydoes it bloom orexert its harmful effects in onesituation and not in another? The answers lie in detailed information about the distributions andactivitiesofphytoplanktonspeciesin relationto theoceanographicandecological processesthatinfluencethem.Thisshouldbecomplementedwithexperimentalresults describinghoweachspeciesresponds to theseenvironmentalfactors, andwithmodels - eitherconceptualormechanistic- thatdescribetheprincipalcontrolsonpopulation dynamicsofthetargetspecies in relation to thephytoplanktoncommunity. Even though it is an enormously daunting task to resolve the complex interac tionsthatdeterminethepopulationdynamicsofaphytoplanktonspeciesincoastalor estuarinewaters, significant progress has been made through the careful work ofin sightful researchers. ThecontributionsofRamon Margalef(e.g. Margalet 1978;Mar galefet aI., 1979) are widely regarded as a seminal influence. Considering life forms ofphytoplankton, that is theirgross morphological and physiological traits, Margalef andcolleaguesdescribedhowfunctionalgroupsofphytoplankton (withrepresentative species identified) could be plottedagainst axes representing nutrient availabilityand the intensity ofturbulence (Margalet 1978). The typical seasonal succession ofphy toplankton, from fast-growing diatoms to motile dinoflagellates, corresponds to the temporal transitionfrom awell-mixed, nutrient-richwatercolumn duringwinterto a nutrient-poor,stratifiedenvironmentlaterintheyear.Thismodelwaslatermodifiedto includea'red-tidesequence', atrajectoryparallelto thetypicalsuccession, butinenvi ronmentswith higherlevelsofnutrients (Figure 1.2). Usingasimilarapproach, Colin Reynolds and colleagues (reviewed briefly by Reynolds and Smayda, 1998; revisited bySmayda,2002) developed'habitatmatrices' thatrelatevariabilityofphytoplankton species composition in lakes to several axes ofvariation, still dominated by nutrients andturbulence. Summarizing efforts to describe variability in phytoplankton communities, Reynolds (2002) concluded that the dynamics of individual species are unpredict able, except on the scale ofdays, and only then ifbased on full knowledge ofinitial distributions. However, he felt that at a higher lever ofgenerality (Le. for functional groupsortrait-basedassociationsofphytoplankton), responsesofphytoplanktoncom munitiestoenvironmentalconditionswouldbeincreasinglypredictable.Akeylinkto predictabilityis thatvariabilityincommunitycomposition is'explicablein retrospect' (Reynolds,2002).Thatis, theconfidencewecanhaveinpredictionsofphytoplankton communitiesunderfuturescenariosdependslargelyonhowwell thesameconceptual modelsexplainhistoricalvariabilityofphytoplankton. As the ecological implications ofmorphological (Karp-Boss et aI., 1996), physi ological and behavioural (Cullen and MacIntyre, 1998) adaptations ofharmful algae become better understood, new and more powerful definitions offunctional groups will emerge, guiding how phytoplankton species should be classified, and which 5 Real-timeCoastalObservingSystems Mucilage-producing cells red-tidesequence \ main succession sequence ) spring Void / winter ~ Nutrientsx turbulence=productionpotential Figure 1.2 The'Mandala' redrawnfromMargalefetal. (1979).Thisdiagram, rich with informationonphytoplanktonsuccession, isextremelyusefulfor developinggeneralizationsaboutrelationshipsbetweenlifeformsof phytoplanktonand hydrographicconditions, particularlyduringseasonal succession.The'red-tidesequence' (developmentofahigh-biomassHAB) canbeviewedasrelatedtoelevatednutrients,independentofchangesin turbulenceregime.Thisconceptualmodelitisnotdirectlyapplicableto real-tiIneobservationandpredictionofHABs.Nonetheless,Margalef's frameworkisacornerstoneofphytoplanktonecologyand itshould supporttherootsofanymodelofseasonalphytoplanktoncommunity dynamicsasinfluencedbylocalconditions. environmental factors must be considered, when trying to predict their dynamics. These more detailed classifications should lead to improvements in predicting the probabilityofHAB occurrence for aparticular location and time, given measured or modelledscenarios ofphysical and chemicalconditions. Moving beyondprobabilities ofoccurrence,thedynamicsofHABsmaybepredictableoverthecourseofdays,given initial data on species distributions from coastal observation systems (Johnsen et aI., 1997;StumpfetaI., 2003). Developmentandtestingofanypredictivemodelwill thus require effective systems for observing the distributions ofphytoplankton, including HABs, in thecontextofcoastalecosystem dynamics. 1.1.3 Observationandmodellinginthe'olddays' ObservationsofHABdynamicshaveseldombeenadequatetodescribethethreestages ofan event: development, maintenanceand decline (TesterandSteidinger, 1997).So, muchofwhatiskno\vnabout HABscomesfrom carefulanalysisoflimiteddata,with much relianceon inference. Comprehensiveobservationsofevents precedingabloom are particularly rare, because the unpredictability that justifies research on harmful algae also precludes the schedulingofcruises to coincidewith HABs. The unpredict ablenatureofHABshas ledto facetious acceptanceofthemaxim thatthebestwayto preventthem is to scheduleamajor research programme tostudythem. 6 Observationandpredictionofharmfulalgalblooms Some ofthe most effective studies ofHABs and other phytoplankton blooms in coastalandestuarinewaterscomefrom regionswhereconditionsaresimilaryeartoyear, andsustainedobservationshaveshown relationships between thepopulationdynamics ofphytoplankton species and hydrographic conditions. Several are described in this volume.AstudybyTylerandSeliger(1978, 1981)isanexcellentexampleofobservation and modelling in the days before autonomous observation systems and three-dimen sionalcoupledmodels. Theapproach theyusedis justas appropriate todayas itwas 25 years ago. The best available sampling techniques were used to characterize the distri butions ofthe target species (Prorocentrum minimum, formerly Prorocentrum mariae lebouriae) in relation to light and hydrographic conditions. In turn, the physiological andbehaviouralresponsesofthesemotilealgaeto thesameenvironmentalfactorswere characterized experimentally. Byconsidering the interaction ofphysiology and behav iourofthephytoplanktonwith theverticalstructureofthewatercolumn andseasonal transportbyestuarinecirculation,selectionforthetargetspecieswasexplainedandthe generalfeaturesofpopulationdynamicspredicted.Allthiswasdonewithlimitedinfor mation. Still, it took 128 ship days during40 cruises over two years (plus hundreds of hourscountingcellsunderamicroscope) toacquirethedatafordescribingthedynamics ofProrocentruminChesapeakeBay(TylerandSeliger, 1978).Subsequently,muchmore has been learned about the dynamics ofProrocentrum minimum in Chesapeake Bay andelsewhere(HeiletaI., 2005), butthevalidityofTylerandSeliger'smultifacetedap proach (observations,experimentation,modelling,validation) has notdiminished. Manyofthe old limitations on coastal ecological research are vanishing. As de scribedin thisvolume, advances inobservation technologyand modelling, supported bygreatlyenhancedcapabilitiesforcommunicationsandcomputing,aretransforming the natureofecological investigation from alabour-intensiveeffort to collectprecious data,interpretedlargelythroughinference,toaprocessinwhichunprecedentedquan tities ofdata and model output must be managed effectively to yield useful informa tion. Still, it is essential to remember that the fundamental principles ofthe research will not change. Species must be identified and their physiological, behavioural and ecological interactions must be considered in the context ofoceanographic processes tounderstandanddescribethepopulationdynamicsofharmfulspeciesasmembersof phytoplanktoncommunities. Insightsfrom the'olddays'ofthetwentiethcenturywill certainlyhelp toguideHAB researchwhenthewidespreadavailabilityofobservations threatens to mask the fundamental need for focused questions about controls on the populationdynamicsofthecausativespecies. 1.2 DETECTION AND PREDICTION FOR MONITORING AND MANAGEMENT OF HABs TheHABproblemisrichwith unansweredquestionsthatwilloccupyscientistsforde cades. Moreimmediately, itrepresents threatstocoastalecosystemsandactivities,and these must be dealtwith now (Malone, 2007 - Chapter 14 this volume). Authorities responsible for environmental protection, economic development and public health must develop and implement plans for the monitoring and management of HABs (AndersonetaI., 2001;Andersen etaI., 2003).These plans may include: • Strategies for monitoring coastal waters for detection ofHABs, including their effects,withan aim to developearlywarningsystems. • Developmentofamodellingsystem forshort-termforecasts ofHAB movements. 7 Real-timeCoastalObservingSystems Integrationofobservations,forecastsandcommunicationsintoanactionplanfor rapidresponsetoHABevents, includingcriteriaforinitiatingstrategicsampling, beachclosures,shellfish bansandcommunication to thepublic. Mitigationstrategies, from directactions to neutralize blooms orminimize their effects (e.g. application ofclay or movement offish cages) to long-term nutrient managementplansorbanson ballastwaterdischarges. Aprogrammeofresearch to predictthelikelihoodofHABs, Le. changes in their frequencyorimpacts, in response to humanactivitiesorclimatechange. Integration ofmonitoring, predictions and communications insupportofpolicy decisions. All aspects ofmonitoring and management require the means to detect, and broad ly to predict, the occurrences and impacts ofHABs on scales from days to decades. These challenges can only be met through fundamental scientific research, but there isnobenefitinconductingthis research independentlyfrom ongoingmonitoringpro grammes. In some jurisdictions, and in the minds ofmany scientists, basic research has beenconsideredto bedistinctfrom routineactivitiessuchasmonitoring.Asaconse quence, monitoringactivitiesmaybespecificallyexcludedfrom fundingprogrammes for research and, in turn, sustained sampling programmes established for research are regularly threatened with cuts or termination, even though they have proved invaluablefordescribinglong-term trends inecosystems (e.g.Tont, 1976; Roemmich and McGowan, 1995; Fromentin and Planque, 1996; Karl et aI., 2001). With the advent ofreal-time coastal observation systems with capabilities for environmental forecasts, the goals ofscientists and coastal managers are aligning, as are the means for attaining these goals: real-time detection ofHABs is essential for earlywarning; prediction of dynamics with forecast models supports rapid response and mitiga tion; and sustained observations (equivalent to monitoring) are required to develop and validate the long-range models ofHAB probabilities needed to develop coastal managementstrategies. Theconclusion is that the research topics ofreal-timedetection, sustainedobser vations and quantitative prediction ofHABs 111ust be integrated with the operational requirements for monitoring and management. To justify the large investment in re search and infrastructure, results from real-tinle coastal observation systems must be accessible, understandable and useful to abroad range ofuser groups. Research must be more closely coordinated with operational oceanography (Chapter 14), which will havetochangewithtimetoserveawiderangeofusers.Thiswill requirerestructuring, not only in the way coastal research is organized, but also in the wayscientists com municate their results- ahealthychallenge. 1.3 CLASSlFlCATION OF HABs FOR OBSERVATION SYSTEMS No observingsystem can provide the oceanographic ideal ofcontinuous andsynoptic measurementsofphysical, chernical and biological propertiesand processes, so efforts mustbe made to match observationsandtheirscales to the HAB phenomenaofinter est (Franks and Keafer, 2003; Srnayda, 2003; Chang and Dickey, 2007 - Chapter 2 this volume). This is no easy task, however. The range ofscales for HABs is inlnlense (Hallegraeff, 2003). Some may be confined to inlets (Seliger et aI., 1970), others are 8 Observationandpredictionofharmfulalgalblooms observedoverlargeexpanses (KahruetaI., 1994). Bloomsmaycomeandgooverdays toweeks, often terminated bywind eventswhich may transport the algae offshore or cause direct mortality from turbulence (but see Smayda, 2002). Other blooms can persistformonths,forexampleKarenia brevisbloomsoffFlorida(TesterandSteiding er, 1997), or even years (Aureoumbra lagunensis in LagunaMadre, Texas; DeYoe and Suttle, 1994).Manyarefound closeto thesurface, notsurprisinglyas discolorationof thewater draws attention. Some ofthese can be traced to subsurface layers, brought near the surface by physical transport, including frontal processes (McMahon et aI., 1998) andvertical mixing (see Section 1.5.1). Othersurface blooms develop through verticalmigrationofphytoplankton (Oliver, 1994;Kamykowski, 1995).Acoarseclas sification ofHABs can be useful as an initial guide to identify relevant scales and appropriate observation strategies for local or regional observation programmes. The classificationissummarizedhere,withsuggestionsforobservationstrategies. 1.3.1 WidespreadHABs Generally, aharmfulalgalbloommusthavewidespreadeffectstoattracttheattention ofthe scientificcommunityand the general public. Discolouredwater, dead fish and noxious foam, scums or aerosols over hundreds ofkilometres ofcoastline make the news, andwithgood reason. The blooms areextensive and the algae areeasilyidenti fied as the cause ofharm. The features that make such blooms noteworthy are also usefulforclassificationandthedesign ofstrategiesfor earlywarning, monitoringand prediction: relativelylarge extent; persistence; and, often but not always, dominance ofthe phytoplankton byonespecies. Three somewhat idealizedcategorieswithin this looseclassificationcan beproposed. 1.3.1.1 Extensive,progressivecoastalblooms Some ofthe most challenging oceanographic and ecological questions are raised by extensive blooms, nearlymonospecificandoften toxic, that appearinacoastalwaters andprogressalongtheshoreline,leavingatrailofshellfishclosures, ravagedfish farms or spoiled beaches. Some examples include blooms ofKarenia brevis in the Gulfof Mexico (Tester and Steidinger, 1997; Stumpfet aI., 2003), Karenia mikimotoi (for merly Gyrodinium aureolum or Gymnodinium mikimotoi) in northern European shelf waters (e.g. Holligan, 1979; DahlandTangen, 1993; Gentien, 1998), the toxicbloom of Chrysochromulinapolylepis in Scandinavian waters in 1988 (Graneli et aI., 1993; Gj0s~teretaI., 2000), bloomsofHeterosigmaintheStraitofGeorgiaandadjacentwa tersin Canada(TaylorandHaigh, 1993), andthedramaticbloomofKareniadigitata in HongKongwaters inApril 1998 (LeeetaI., 2007- Chapter 18thisvolume). Hypothesesaboutbloomdynamicsfocus ontheprocessesofinitiation, transport andinteractionsofpopulationswithsurfacecirculation (TesterandSteidinger, 1997). Assessmentofimpacts requires information on transmission ofharm (e.g. decaylead ingtoanoxia, directcontact, toxintransferthrough ingestion, productionofaerosols), how these processes relate to the distributions ofharmful algae, and environmental influencesontheproductionofnoxiouseffects. Many environmental properties must be measured for effective earlywarning, monitoringandprediction.Whenconditionspermit,remotesensingofoceancolour and sometimes sea surface temperature from satellites and aircraft can provide key information on distributions and transport (Stumpf et aI., 2003; Ruddick et aI., 2007 - Chapter 9 this volume), especially when supplemented by observation 9 Real-timeCoastalObservingSystems networks that include direct sampling (Johnsen et al., 1997; Tangen, 1997). Even jfsurface distributions ofdeveloped blooms are resolved with remote sensing, early stages andsubsurfacedistributions mustbedescribed byothermeans. In particular, vertical distributions of phytoplankton should be well resolved because the inter action ofswimming, sinking or floating with frontal features (Franks, 1997), ag gregation ofseed populations in subsurface layers near the pycnocline (McMahon et al., 1998), and changes ofbehaviour in mixed waters landward ofa front (Dahl and Tangen, 1993; Gentien, 1998), possibly associated with nutrition (Cullen and MacIntyre, 1998), all may be importantin initiation, maintenance and transportof extensive, progressive, coastal blooms. Consequently, for early warning and moni toring, observation systems must resolve vertical distributions ofphytoplankton in relationto temperature,salinityandcurrents,andtheymusthavethemeansto iden tify target species in situ. Nutrient availability can influence toxicity (Bates, 1998; Cembella, 1998) and depletion ofnutrients can terminate a bloom. So, for effective monitoringand modelling, the nutrient regimeshould also beassessed. Progressive coastal blooms move with coastal currents and can appear or disap pear on the timescale ofdays. Effective monitoring thus requires nearly continuous measurements, and mitigation responses (suchas the movementofaquaculturecages) requirecommunicationsinnearrealtime(Tangen, 1997).Strategiesfor management, suchascontrolsoncoastalnutrientloadingorsiteselectionforaquaculture,dependon longtimeseriesofobservations to determinetherelationshipsbetweenenvironmental variability, humaninfluences,bloomoccurrencesandtheirimpacts.Sustaineddeploy mentofreal-timeobservationsystemsis thusideallysuitedforobservationandpredic tionofextensive, progressivecoastalblooms. 1.3.1.2 Extensivebloomsinopenwaters It has long been recognized that phytoplankton blooms in open waters are part ofthe natural ecologyofthe oceans. The phenomenologyofsome, such as the vernal diatom bloomintemperatewaters(Sverdrup, 1953),surfaceaggregationofTrichodesmiumduring calmperiods(Caponeetal., 1998),Phaeocystisanddiatombloomsassociatedwithreced ingiceedges(Lancelotetal., 1998)andthegreenwatersofupwellingsystems(Barberand Smith, 1981) are fairly well understood. The root causes ofdramatic expanses ofmilky water from coccolithophores are open to informed speculation (e.g. Olson and Strom, 2002) andexaminationthroughnumericalmodelling(Mericoetal., 2004).Satelliteim agery4revealsmanyotherbloomsinopenwatersthatwillremaincuriositiesuntiltheyare studiedfurther. Inthecontextofthischapter,interestisfocusedonharmfulorpotentially harmfulbloomsthatoccurinopenwatersinsemi-enclosedseasornearcoasts,wherethey can influencecoastal ecosystems and be affected by terrestrial inputs offresh waterand nutrients.TheBaltic,NorthSeaandBohai(China)areexemplary. Itserveslittlepurpose toapplythisclassificationstrictly;extensivebloomsinopenwatersaregroupedsothepo tential forcingfunctions- climatechangeandnutrientsources- can bediscussedalong withstrategiesforobservingandpredictingecologicalresponsestotheseinfluences. OpenwaterHABscancauseproblemswhentheyimpingeonthecoast,delivering scums, foams ortoxicity. Forexample,summerbloomsofnitrogen-fixingcyanobacte riaintheBalticSeaarecommon (Sellner, 1997).ThehepatotoxicNodulariaspumigena is conspicuous; during the latter stages ofa bloom, filaments form highly visible ag gregates at thesurface that can be detected from space (Kahru etal., 1994). Nitrogen 4Forexample:http://visibleearthnasa.gov/ 10

Description:
describing how each species responds to these environmental factors, and with .. dure ofobservation, analysis and improved observations and Tester, 1997) and shellfish (e.g. Tracey, 1988), and competition .. termination ofchlorophyll concentration, measurement of turbidity, and enumeration. 20
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.